Background: Radiographs of the sacroiliac joints are commonly used for the diagnosis and classification of axial spondyloarthritis. The aim of this study was to develop and validate an artificial neural network for the detection of definite radiographic sacroiliitis as a manifestation of axial spondyloarthritis (axSpA). Methods: Conventional radiographs of the sacroiliac joints obtained in two independent studies of patients with axSpA were used. The first cohort comprised 1553 radiographs and was split into training (n = 1324) and validation (n = 229) sets. The second cohort comprised 458 radiographs and was used as an independent test dataset. All radiographs were assessed in a central reading session, and the final decision on the presence or absence of definite radiographic sacroiliitis was used as a reference. The performance of the neural network was evaluated by calculating areas under the receiver operating characteristic curves (AUCs) as well as sensitivity and specificity. Cohen’s kappa and the absolute agreement were used to assess the agreement between the neural network and the human readers. Results: The neural network achieved an excellent performance in the detection of definite radiographic sacroiliitis with an AUC of 0.97 and 0.94 for the validation and test datasets, respectively. Sensitivity and specificity for the cut-off weighting both measurements equally were 88% and 95% for the validation and 92% and 81% for the test set. The Cohen’s kappa between the neural network and the reference judgements were 0.79 and 0.72 for the validation and test sets with an absolute agreement of 90% and 88%, respectively. Conclusion: Deep artificial neural networks enable the accurate detection of definite radiographic sacroiliitis relevant for the diagnosis and classification of axSpA.

A deep learning tool was able to detect active inflammatory and structural changes indicative of axial spondyloarthritis at MRI of sacroiliac joints. Background MRI is frequently used for early diagnosis of axial spondyloarthritis (axSpA). However, evaluation is time-consuming and requires profound expertise because noninflammatory degenerative ...

Purpose To examine whether incorporating anatomy-centred deep learning can improve generalisability and enable prediction of disease progression. Methods This retrospective multicentre study included conventional pelvic radiographs of four different patient cohorts focusing on axial spondyloarthritis collected at university and community hospitals. The first cohort, which consisted of 1483 radiographs, was split into training (n=1261) and validation (n=222) sets. The other cohorts comprising 436, 340 and 163 patients, respectively, were used as independent test datasets. For the second cohort, follow-up data of 311 patients was used to examine progression prediction capabilities. Two neural networks were trained, one on images cropped to the bounding box of the sacroiliac joints (anatomy-centred) and the other one on full radiographs. The performance of the models was compared using the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity and specificity. Results On the three test datasets, the standard model achieved AUC scores of 0.853, 0.817, 0.947, with an accuracy of 0.770, 0.724, 0.850. Whereas the anatomy-centred model achieved AUC scores of 0.899, 0.846, 0.957, with an accuracy of 0.821, 0.744, 0.906, respectively. The patients who were identified as high risk by the anatomy-centred model had an OR of 2.16 (95% CI 1.19, 3.86) for having progression of radiographic sacroiliitis within 2 years. Conclusion Anatomy-centred deep learning can improve the generalisability of models in detecting radiographic sacroiliitis. The model is published as fully open source alongside this study.

BACKGROUND: Radiofrequency ablation (RFA) is an important treatment option for hepatic tumors and metastases. Post-ablation recurrence rates are reported up to 36.5 percent and seem to depend on tumor size, intrahepatic localization of tumors and adj

To evaluate the diagnostic value of SpA parameters and their combination for the diagnosis of axial SpA in patients with an a priori different probability of the diagnosis. A total of 361 patients with chronic back pain and suspicion of axial SpA (181 referred by primary care physicians or orthopaedists, 180 recruited via an online screening tool) received a structured rheumatologic examination, which resulted into a diagnosis or exclusion of axial SpA. The prevalence of axial SpA indicating the pre-test probability was 40% in the physician-referred subgroup and 20% in the online screening subgroup. Sensitivities, specificities and likelihood ratios for SpA features were determined in both subgroups and the respective post-test probabilities of axial SpA were calculated. The relative diagnostic value of single SpA features varied substantially between the groups with different referral pathways. For instance, HLA-B27 positivity increased the probability of the presence of axial SpA by 35% to 55% in online-screened patients and by 22% to 62% in physician-referred patients. The absence of HLA-B27 resulted in a sharp decrease in the probability of the presence of axial SpA in physician-referred patients (from 40% to 6%). This decrease was less sharp in the online screening group (from 20% to 10%). These differences were especially relevant in patients with a small number (one to two) of positive SpA features. The diagnostic value of SpA features varies in different patient populations, which should be considered in the diagnostic approach.